A rule-based classifier uses a set of If - Then rules for classification. An If - Then rule is an expression of the form
If condition Then conclusion.
The If part is known as rule antecedent or precondition and Then part is known as rule consequent. The rule antecedent contains attribute test condition and rule consequent contains class prediction.
If rule antecedents are satisfied, then the rule covers the data in the dataset.
The coverage and accuracy of a rule is assessed by
coverage = ncovers/|D|
accuracy = ncorrect/ncovers
where
ncovers - no of data covered by rule
ncorrect - no of data classified correctly by rule
D - no of data in the dataset.
If a rule is satisfied by a data in the dataset, then the rule is said to be triggered. IF only one rule is triggered, then the rule classifies the data, which referred as rule firing. If more than one rule is triggered, then conflict resolution strategy to be applied to classify the data.
The size ordering and rule ordering are the conflict resolution strategies.
The size ordering strategy assigns the highest priority to the triggering rule with more attribute tests.
The rule ordering may be class based or rule based. The classes are sorted in decreasing order of prevalence.
The rules are organized into one long priority list, according to some measure of rule quality, such as accuracy, coverage, or size or based on advice from domain experts in the rule based ordering. When rule ordering is used, the rule set is known as a decision list.
Most rule-based classification systems use a class-based rule-ordering strategy.